Presentation 2008-07-18
Speaker diarization for meetings by integrating speech presence probability estimation and time-frequency domain direction of arrival estimation
Shoko ARAKI, Masakiyo FUJIMOTO, Kentaro ISHIZUKA, Tomohiro NAKATANI, Hiroshi SAWADA, Shoji MAKINO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper presents a meeting diarization system that estimates who spoke when in a meeting. Our proposed system is realized by using a noise robust voice activity detector (VAD), a direction of arrival (DOA) estimator, and a DOA classifier. This paper proposes two methods for improving diarization performance. As the first proposal, we employ a DOA at each time-frequency slot (TFDOA) so that multiple DOAs can be estimated at a frame when multiple speakers speak simultaneously. The second proposal is to integrate VAD and DOA in a probabilistic way. This paper reports how such proposals improve diarization performance for real meetings/conversations.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) diarization / voice activity detector / direction of arrival
Paper # EA2008-40
Date of Issue

Conference Information
Committee EA
Conference Date 2008/7/11(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Engineering Acoustics (EA)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Speaker diarization for meetings by integrating speech presence probability estimation and time-frequency domain direction of arrival estimation
Sub Title (in English)
Keyword(1) diarization
Keyword(2) voice activity detector
Keyword(3) direction of arrival
1st Author's Name Shoko ARAKI
1st Author's Affiliation NTT Communication Science Laboratories, NTT Corporation()
2nd Author's Name Masakiyo FUJIMOTO
2nd Author's Affiliation NTT Communication Science Laboratories, NTT Corporation
3rd Author's Name Kentaro ISHIZUKA
3rd Author's Affiliation NTT Communication Science Laboratories, NTT Corporation
4th Author's Name Tomohiro NAKATANI
4th Author's Affiliation NTT Communication Science Laboratories, NTT Corporation
5th Author's Name Hiroshi SAWADA
5th Author's Affiliation NTT Communication Science Laboratories, NTT Corporation
6th Author's Name Shoji MAKINO
6th Author's Affiliation NTT Communication Science Laboratories, NTT Corporation
Date 2008-07-18
Paper # EA2008-40
Volume (vol) vol.108
Number (no) 143
Page pp.pp.-
#Pages 6
Date of Issue